Referrals are the hardest channel to measure and the most important to understand. There is no click, no campaign, and no analytics report. Which is why most practices know their referrals matter and cannot say who sends them.
Ask, and record the answer
Everything here rests on one habit: ask every new patient how they came to you, and record the answer in a structured field.
Not a free-text note. A defined list you can count:
- Referred by a clinician, and which one.
- Insurance directory, and which plan.
- Search engine.
- An AI assistant.
- A specific vertical directory.
- Word of mouth from a friend or family member.
- Returning patient.
Two disciplines make this work: ask at intake every time, without exception, and capture the specific referrer name rather than just clinician referral. A count of referrals is mildly interesting; a count by referring clinician is actionable.
What to track per referrer
Build a simple record, one row per referring clinician:
- Referrals received, by month.
- How many were reached and scheduled.
- How many attended an evaluation.
- How many proceeded to treatment.
- Date of first referral and date of most recent.
- Whether you have sent them a consultation note each time.
The two dates are the most valuable and the least recorded. A referrer who sent four patients last year and none in five months is the single clearest signal in the whole dataset, and without the recency date you will not notice.
The metrics that matter
- Active referrers. How many distinct clinicians referred in the last quarter. This measures the health of the network rather than its size.
- Concentration. What share of referrals comes from your top three sources. High concentration is a real business risk: one retirement can remove a third of your volume.
- Lapsed referrers. Sources who used to refer and have stopped.
- Referral to evaluation rate. If referrals arrive but do not attend, the problem is your intake, not your referrers.
- Evaluation to treatment rate. A low rate may mean referrers are sending unsuitable patients, which means your one-pager needs clearer exclusions.
- New referrers per quarter. Whether the network is growing.
Reading the results
Each pattern implies a specific action:
- A referrer stopped. Find out why. Usually they did not hear back, their patient waited too long, or they did not know whether the patient was returned to them.
- Referrals not converting to evaluations. An intake failure. Check contact speed and attempt count.
- Evaluations not converting to treatment. Either referral criteria are unclear, or cost and insurance are stopping people. Both are addressable.
- Heavy concentration. Deliberately broaden the network before you are forced to.
- Payer directory patients appearing. Evidence that chapter eight accuracy work is paying, and the only way you will ever see it.
What to do with it
- Thank active referrers, appropriately and without anything of value changing hands.
- Reconnect with lapsed referrers, leading with what you send back rather than with a pitch.
- Prioritise your time on the sources actually producing, not the ones you assume are.
- Fix the intake and communication failures the data exposes, since those are cheaper than finding new referrers.
The compliance boundary
Measuring referral sources is normal practice management. What you must not do is convert appreciation into anything of value. Do not build reward schemes, gifts of significance, fee arrangements, or any inducement tied to referral volume, because this raises serious issues under healthcare fraud and abuse laws. Take advice before any financial arrangement with a referral source.
Track, thank, communicate well, and let good clinical service do the work. Chapter nine covers measurement across every channel.